Classifications

• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T13/00—Animation
  • G06T13/80—2D [Two Dimensional] animation, e.g. using sprites
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T13/00—Animation
  • G06T13/20—3D [Three Dimensional] animation
  • G06T13/40—3D [Three Dimensional] animation of characters, e.g. humans, animals or virtual beings
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2210/00—Indexing scheme for image generation or computer graphics
  • G06T2210/21—Collision detection, intersection

Definitions

• This invention relates to a method for animating characters, and in particular for animating large groups of characters, also called crowds.
• Crowd simulation is widely used in games, movies, urban planning and emergency evacuation simulation.
• a group of characters, or crowd is a collection of people/characters that stick together, have the same goal and behave similarly. Group behaviour is very common in our daily life and has been widely studied. If there is strong grouping behaviour in a crowd, it is known to simulate it by first planning the path for each group as a whole, and then planning the crowd.
• information about at least the computed new velocity will then be stored in cells of the terrain map or roadmap, and it will influence at least the next character that enters the cell.
• the roadmap influences individual behaviour of at least the next character, not the group behaviour in general.
• the information is not related to fixed terrain obstacles, as used for path planning. Instead, the information stored in the terrain map is only an intermediate result, and it is used as a
• the invention deals with how to size the grid cells, how to use it for collision avoiding, how to organize tracing information, and how to judge the effectiveness of the information.
• a device for determining for a character in a group at least one of a moving
• a moving velocity comprising a memory for storing cells of a terrain map, first data read means for reading tracing information from a current cell in a terrain map on which the character is located, a first collision detector for determining, based on the tracing information from the first data read means, whether or not collision avoidance is needed, and for notifying the determining result, a first position calculator for advancing the character according to its previous direction and previous velocity if the first calculator notifies that collision avoidance is not needed, a first processor for searching, if the first calculator notifies that collision avoidance is needed, for at least one of a matching direction and
• a second processor for computing a time-to- collision based on the found matching direction and/or velocity, a comparator for comparing the computed time-to- collision with a predetermined range, and a second position calculator for advancing the character according to the found matching direction and/or velocity, and for performing a collision avoiding manoeuvre and updating the tracing information in the current terrain cell if no matching direction and/or velocity is found, or if the time-to- collision is outside the predetermined range.
• a given terrain is cut into cells of a
• the size is determined according to various parameters, e.g. typical velocity (see below). For a
• the entry direction is determined (e.g. one of north N, east E, south S, west W, or one of north N, north-east NE, east E, south-east SE etc.), and it is determined what the direction and/or velocity decision was for previous characters that entered from the same direction. Then it may be checked if certain
• Reciprocal Velocity Obstacle as described in "Realtime rendering of densely populated urban environments", F. Tecchia, Y. Chrysanthou, Proceedings of the Eurographics Workshop on Rendering Techniques, 2000, is a new technique dealing with collision avoidance.
• RVO Reciprocal Velocity Obstacle
• there is an RVO area (or range) for each of the characters on the velocity plane which indicates the velocities that the characters ought not to choose, so that collisions can be prevented. If a character chooses any velocity out of the RVO area, it is guaranteed to avoid the collision with the other character.
• the RVO technique also assumes that each of the characters adopts the same strategy to avoid collision. In other words, the characters share the responsibility and take action accordingly to avoid any collision .
• Fig.3 a portion of a terrain grid
• Fig.4 a grid portion and tracing information
• Fig.5 a block diagram of a device for determining at least one of a moving direction and a moving velocity for a character in a group.
• Fig.l shows a multi -character scenario RVO.
• one character 10 should choose a velocity which lies out of the combination of the RVOs of all the neighbouring characters.
• the neighbouring characters 11 have individual moving directions and speeds 12. However, in a very crowded
• Vi' is the new velocity
• Vi pref is the preference velocity
• tCi(Vi') is the estimated time to collision
• wi is the weight to balance between time to collision and velocity diversion.
• the proposed method reduces the complexity of crowd
• the plane e.g. a 2D plane
• the resolution of the grid is determined by the preference velocity and simulation time step.
• the tracing information including the entry angle, the new velocity and the time to collide, is stored in each cell of the grid.
• the following character searches for the tracing information which matches its entry angle (i.e. that has substantially the same entry angle) , and then computes the time to collide based on the new velocity. If the time to collide for the current character is comparable with that stored in the cell (at least if both are equal) , the
• Fig.2 shows the steps of the proposed simulation framework. Detailed explanation for each step is given below.
• Fig.3 shows a portion of a terrain grid with a diameter d, which is called resolution.
• the resolution of the grid is determined by the principle that for each time step, the character should walk with
• Fig.4 shows a grid portion and tracing information.
• the tracing information stored in a grid cell includes at least four sets of data, each of which relates to a range of entry direction Ve of the character.
• the new velocity V ne w and the time-to-collide (TTC) based on this new velocity are included and presented.
• TTC time-to-collide
• a simulation process is performed. If collision avoiding is needed, a first module PI performs searching for a match in a search step 22. That is, whenever it is determined 21 that the collision avoiding manoeuvre is needed, we search 22 the cell being occupied by the character for determining 23 if tracing information exists that matches the entry direction of this character. If the tracing information does not exist, a regular
• collision avoiding manoeuvre is conducted 27, which is known and usually highly complex.
• a second module P2 computes a time-to- collision (TTC) .
• TTC time-to- collision
• TTC is the time it takes for the first collision to happen. If it is determined 25 that no collision will happen, TTC is sufficiently large.
• This computed TTC should be subjected to TTC com P uted > ma (TTC tracing - k*At, At) (3) wherein TTC trac:Ln9 is the tracing information stored in the cell. Otherwise the match fails and a regular collision avoiding manoeuvre will be conducted 27.
• the constant k indicates a number of simulation time steps and should be chosen properly to indicate a tolerance of mismatch.
• a third module P3 performs 27 a regular collision avoiding manoeuvre. If no tracing information exists in a cell, or if available tracing information does not match well, the regular collision avoiding maneuver will be conducted in order to select a new velocity for the current character.
• a fourth module P4 updates 28 the tracing information in the terrain cell. Whenever the regular collision avoiding
• the new position is stored, either in the terrain map or in a separate memory.
• the invention concerns a device for determining at least one of a moving direction and a moving velocity for a character in a group.
• Fig.5 shows a block diagram of such device.
• the device for determining for a character in a group at least one of a moving direction and a moving velocity comprises
• first data read means 52 e.g. memory accessing unit
• first collision detector 53 for determining, based on the tracing information from the first data read means, whether or not collision
• a first position calculator 54 for advancing the character according to its previous direction and previous velocity if the first calculator notifies that collision avoidance is not needed nca, a first processor 55 for searching, if the first calculator notifies that
• a second processor 56 for computing a time-to-collision TTC based on the found matching direction and/or velocity
• a comparator 57 for comparing the computed time-to-collision TTC with a
• a second position calculator 58 for advancing the character according to the found matching direction and/or velocity, and for performing a collision avoiding manoeuvre and updating the tracing information in the current terrain cell if no matching direction and/or velocity is found, or if the time-to-collision is outside the predetermined range .
• the invention can be used in different application scenarios. Generally, the method reduces the complexity of crowd
• Tracing information stored in a terrain cell is used as guidance for a current character that has a similar entry direction as another character that was
• This method is based on the assumption that the characters in a group that follow others will face a situation similar to that of previous characters. This is especially true when two groups of characters cross each other.
• the method of the invention can be used to determine motion direction and velocity of individual characters in groups that
• the method can be used in any crowd simulation, e.g. fighting scene in movies, traffic simulation, evacuation simulation, etc., as long as the assumption mentioned above is justified.
• the proposed method is an effective way to simulating very large crowd (thousands of characters, or more) .
• d i.e. the indicator of resolution of the grid
• k k in the TTC formula
• the present invention is usable for simulating a crowd e.g. by first planning the path and velocity for each group and then adjusting the velocity for each character inside the group. Steps are
• each group is treated as a single entity (i.e. a group heading direction is determined for each group) .

Landscapes

• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Engineering & Computer Science (AREA)
• Theoretical Computer Science (AREA)
• Processing Or Creating Images (AREA)

Priority Applications (5)

Applications Claiming Priority (2)

Publications (1)

Family

ID=46145388

Family Applications (1)

Country Status (5)

Cited By (2)

Families Citing this family (10)

Citations (3)

Family Cites Families (14)

• 2011
  • 2011-11-28 KR KR1020137013299A patent/KR20130133778A/ko not_active Withdrawn
  • 2011-11-28 US US13/988,338 patent/US9460540B2/en not_active Expired - Fee Related
  • 2011-11-28 WO PCT/CN2011/083077 patent/WO2012069022A1/en not_active Ceased
  • 2011-11-28 JP JP2013540230A patent/JP5905481B2/ja active Active
  • 2011-11-28 EP EP11843211.1A patent/EP2643817A4/en not_active Withdrawn

Patent Citations (3)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events